Hypomorphic Mutation of PDK1 Suppresses Tumorigenesis in PTEN+/− Mice Jose R. Bayascas, Nick R. Leslie, Ramon Parsons, Stewart Fleming, Dario R. Alessi  Current Biology  Volume 15, Issue 20, Pages 1839-1846 (October 2005) DOI: 10.1016/j.cub.2005.08.066 Copyright © 2005 Elsevier Ltd Terms and Conditions

Figure 1 Generation and Analysis of PDK1 Hypomorphic PTEN+/− Mice (A) Diagram illustrating the positions of the exons 2 to 5 of the PDK1 gene [15] and 4 to 6 of the PTEN gene [11], depicting the different alleles that we have utilized. The black boxes represent exons, the continuous lines introns, and triangles represent CRE loxP excision site and NEO the neomycin resistance gene cassette. (B) Breeding strategy employed to generate the control PDK1+/flPTEN+/− and experimental PDK1−/flPTEN+/− mice used in this study. The number and percentage of each genotype obtained after weaning are indicated. An asterisk indicates lower frequency than expected of PDK1−/flPTEN+/+ obtained is statistically significant (χ2 Test and Fisher’s Exact Test p < 0.05). (C) The mean body weight of the control PDK1+/flPTEN+/− and experimental PDK1−/flPTEN+/− mice at the indicated age, where n corresponds to the number of mice in each group. Values represent the mean ± SEM for each data point. (D and E) Liver extracts for the indicated mice were prepared and PDK1 (D) or PTEN (E) were immunoprecipitated and assayed. The results shown are the average ± SEM corresponding to samples derived from three independent mice assayed in triplicate. Samples from three different mice of each genotype were also immunoblotted with the indicated antibodies. The double asterisks in (C) and (D) indicates p < 0.005 as obtained by the Student’s t test. Current Biology 2005 15, 1839-1846DOI: (10.1016/j.cub.2005.08.066) Copyright © 2005 Elsevier Ltd Terms and Conditions

Figure 2 Analysis of Survival and Tumor Formation (A) Mice were maintained under standard husbandry conditions and the percentage of surviving mice is indicated. (B) The percentage of mice with visible tumors is indicated. For mice up to 12 months of age, these tumors only comprise those that were externally visible. At 14–15 months of age when the study was terminated, tumors that were visible internally after dissection are also included in the analysis. n corresponds to the number of mice of each genotype. Arrows in (B) indicate the onset of tumor development for each group of mice. An asterisk indicates p < 0.05, and a double asterisk, p < 0.005 obtained by both the χ2 Test and the Fisher’s Exact Test. Current Biology 2005 15, 1839-1846DOI: (10.1016/j.cub.2005.08.066) Copyright © 2005 Elsevier Ltd Terms and Conditions

Figure 3 Histopathological Analysis of PDK1+/flPTEN+/− and PDK1−/flPTEN+/− Mice The pathology of the several tumor types shows no differences between those occurring in PDK1+/fl PTEN+/− and those occurring in PDK1−/flPTEN+/−. (A and B) Non Hodgkin’s lymphomas were of a diffuse predominantly small cell type with some plasmacytoid differentiation. (D and E) Prostatic carcinomas showed a cribriform architecture on occasion with comedo-type necrosis and invasion of the surrounding stroma. (F and G) Phaeochromocytomas were composed of large polygonal amphophilic cells with granular cytoplasm and a rich vascular supply. (H and I) The breast carcinomas were of an invasive ductal type eliciting a desmoplastic reaction. (J) Testicular tumors, in this example a Leydig cell tumor, were only seen in the control animals as were endometrial carcinomas (C); preneoplastic lesions of prostate amounting to high grade PIN (K and L) were seen in both groups of animals; as was atypical endometrial hyperplasia (M and N). Haematoxylin and eosin, original magnification ×120. (A), (C), (D), (F), (H), (J), (K), and (M), PDK1+/fl PTEN+/−; (B), (E), (G), (I), (L), and (N), PDK1−/fl PTEN+/−. Current Biology 2005 15, 1839-1846DOI: (10.1016/j.cub.2005.08.066) Copyright © 2005 Elsevier Ltd Terms and Conditions

Figure 4 Immunohistochemistry Analysis of Tumors from PDK1+/flPTEN+/− and PDK1−/flPTEN+/− Mice Sections derived from prostate carcinoma (A) and breast adenocarcinoma (B) of PDK1+/flPTEN+/− (panels I III, and V) and PDK1−/flPTEN+/− (panels II, IV, and VI) tumors were probed with the indicated antibody, counterstained with haematoxylin, and visualized with a magnification of ×160. In panels I and II, cytoplasmic reactivity of FOXO1 was mainly observed in tumor cells and to a lesser extent in fibroblasts of both breast adenocarcinomas. In panels III and IV, phospho-S6 protein phosphorylated at Ser235/Ser236 was seen by strong cytoplasmic reactivity in all tumor samples. In the breast adenocarcinoma, both tumor cells and the adjacent reactive stromal fibroblasts were observed to be stained. In panels V and VI, Ki67 staining showed that in both types of carcinoma about 25% of the tumor cells are in cell cycle with reactivity also noted in the stromal fibroblasts of the breast adenocarcinoma. Current Biology 2005 15, 1839-1846DOI: (10.1016/j.cub.2005.08.066) Copyright © 2005 Elsevier Ltd Terms and Conditions